Preventing Unnecessary Groundings in the Lifted Dynamic Junction Tree Algorithm

TL;DR

This paper extends the lifted dynamic junction tree algorithm to identify and prevent unnecessary groundings, significantly improving the efficiency of answering temporal probabilistic queries.

Contribution

The authors introduce methods to detect and delay eliminations that cause groundings, enhancing LDJT's performance in probabilistic relational temporal models.

Findings

Extended LDJT answers queries orders of magnitude faster.

Successfully prevents unnecessary groundings during inference.

Improves efficiency without sacrificing accuracy.

Abstract

The lifted dynamic junction tree algorithm (LDJT) efficiently answers filtering and prediction queries for probabilistic relational temporal models by building and then reusing a first-order cluster representation of a knowledge base for multiple queries and time steps. Unfortunately, a non-ideal elimination order can lead to groundings even though a lifted run is possible for a model. We extend LDJT (i) to identify unnecessary groundings while proceeding in time and (ii) to prevent groundings by delaying eliminations through changes in a temporal first-order cluster representation. The extended version of LDJT answers multiple temporal queries orders of magnitude faster than the original version.